Dear James,
I must say I have no idea what is going on here. Like you I would have predicted an increase in FA, if anything.
What do you see if you run a movie of the data (prior to any dtifit) corrected with and without slice-to-vol?
Jesper
> On 12 Nov 2018, at 19:45, James Kennedy <[log in to unmask]> wrote:
>
> Hello,
>
> I'm testing out the new s2v option in FSL 5.0.11's eddy_cuda as part of a pipeline with AutoPtx designed to extract tract size and average FA, L1, L2, L3, and MD. Adding " --mpoder=6 --slspec=slspec.txt --s2v_niter=5 --s2v_lambda=1 --s2v_interp=trilinear" drops my average FA to 58% of my initial analysis, with my average L1 increasing 48%, L2 increasing 94%, L3 increasing 135%, and MD increasing 78%. The tracts created by AutoPtx are essentially identical so it's not an issue of different regions being identified. Changing s2v_niter to 5 and s2v_lambda to 10 gave similar results. The diffusion data is 102 directions, 81 slices with a multiband factor of 3. Running dtifit shows this to be an issue for the whole brain in the attached images (range 0 to 1 for both). I was expecting my FA to increase, if anything. Any idea what's happening here?
>
> Thank you,
> James Kennedy
> Washington University
>
> ########################################################################
>
> To unsubscribe from the FSL list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
> <No_s2v.png><s2v.png>
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|